• Selected
• Entire List

Abstract:
We explore the link between the overnight fed funds rate, which is actively targeted by the Federal Reserve, and longer-maturity term fed funds rates. We develop a term-structure model which explicitly accounts for interest rate targeting and for the predictability of future target changes. The model is able to replicate some qualitative features of the dynamic behavior of deviations of short-term rates from the target.

fed funds rates, expectation hypothesis, autocovariance functions

Abstract:
We find that in 1989-1996, when U.S. monetary policy tightly targeted overnight fed funds rates, the volatility and persistence of spreads between target and term fed funds levels were larger for longer-maturity loans. We show that such patterns are consistent with an expectational model where target revisions are infrequent and predictable. In our model, the (autoco-) variance of the spreads of term fed funds rates from the target increases with maturity because longer-term rates are more heavily influenced by persistent expectations of future target changes.

Abstract:
A characteristic feature of U.S. monetary policy has been the active targeting of the overnight fed funds rate by the Federal Reserve. We show that during the 1989-1996 period, in spite of the effective targeting of the overnight fed funds rate, term fed funds rates displayed volatile and persistent spreads from the target. Moreover, the volatility and persistence of these spreads increase with the maturity of the loan. This behavior is consistent with an expectational model of short-term rates which accounts for interest rate targeting with predictable and infrequent revisions of the target, on a daily time scale. Our model successfully replicates the stylized fact that the (autoco-)variance of the spreads of term fed funds rates from the target increases with maturity, because longer-term rates reflect more heavily persistent expectations of the next target change.

Abstract:
A characteristic feature of U.S. monetary policy has been the active targeting of the overnight fed funds rate by the Federal Reserve. We show that during the 1989-1996 period, in spite of the effective targeting of the overnight fed funds rate, term fed funds rates displayed volatile and persistent spreads from the target. Moreover, the volatility and persistence of these spreads increase with the maturity of the loan. This behavior is consistent with an expectational model of short-term rates which accounts for interest rate targeting with predictable and infrequent revisions of the target, on a daily time scale. Our model successfully replicates the stylized fact that the (autoco-)variance of the spreads of term fed funds rates from the target increases with maturity, because longer-term rates reflect more heavily persistent expectations of the next target change.

Abstract:
This paper examines newly-available intra-day data from the inter-dealer government bond market to investigate the effects of economic-news announcements on prices, trading volume, and bid-ask spreads. The use of intra-day data together with data on market expectations allows us to obtain new and different results relative to previous studies. We find a total of seventeen economic announcements to have a significant impact on the price of at least one of the following instruments: a three-month bill, a two- and ten-year note, and a thirty-year bond. Ten of them significantly affect all note and bond prices. For announcements that have a significant impact on prices, the impact occurs within one minute after the announcement. Interestingly, only three announcements affect the bill price. This suggests that at least two factors of uncertainty are needed to model the yield curve. For the ten-year note we find a strong association between announcements and trading volume. Economic announcements have less effect on trading volume for the three-month bill, although changes in monetary policy lead to an average trading volume up to nine times higher than at non-announcement times. Bid-ask spreads widen immediately after most economic announcements, but then return to normal levels within 5 to 15 minutes. For almost all announcements, volatility is significantly higher after the release, especially for the announcements that significantly affect prices.

Abstract:
In one-factor models, such as Cox, Ingersoll, and Ross (1985) or Vasicek (1977), the conditional mean of the instantaneous rate changes with its current level. This paper gathers evidence that the conditional mean of the one- month rate explains variations in bond yields of different maturities, even after controlling for the effect of the current level of the one-month rate. This suggests the presence of a second factor driving the conditional mean, other than the level of the one-month rate: we refer to this second factor as the central tendency. The above idea is captured in a two-factor model of the term structure where the instantaneous rate fluctuates around a stochastic central tendency. We then build a proxy for the central- tendency factor based on the information contained in the term structure of interest rates. We use the proxy to estimate the process for the one-month rate, and find the central-tendency proxy to be significant in explaining the conditional mean of the one-month rate.

Abstract:
This paper examines portfolio choice, trading behavior, and realized rates of return of more than seven thousand 401(k) retirement accounts during the April 1994-August 1998 time period. The evidence on equity allocations is indicative of prudent behavior: on average our investors hold 40% of their 401(k) portfolios in stocks. In addition, patterns of stock allocations by marital status, age, and earnings are broadly consistent with the implications of normative models: stock allocations are higher for married investors, for younger investors, and for investors with higher earnings. The evidence on trading activity indicates very limited portfolio re-shuffling, which stands in sharp contrast to existing evidence from discount brokerage accounts: 70% of the plan participants do not rebalance their portfolio more than once, average re-balancing frequency is one trade every 33 months, and average monthly turnover is in the order of 2%. This evidence is consistent with the implications of models of optimal portfolio choice with realistic transaction costs.

Social Security, Retirement, 401(k), Portfolio

Abstract:
Starting from the cross-sectional aggregation of marginal utilities, rather than intertemporal marginal rates of substitution, this paper develops a new heterogeneous-agent pricing kernel. A closed-form version of the new kernel depends on changes in the cross-sectional variance of log consumption, rather than the variance of changes in log consumption, as in Constantinides and Duffie (1996). We implement the new kernel on household consumption data from the Consumer Expenditure Survey (CEX). At monthly and quarterly horizons, the new kernel reconciles the premium on U.S. equities with the consumption of asset holders, for reasonable values of relative risk aversion. The new kernel also fares better than kernels based on the aggregation of intertemporal marginal rates of substitution in explaining the cross-sectional variation of risk premia on stocks and bonds. Portfolios mimicking changes in the cross-sectional variance of log consumption for asset holders command significant negative Sharpe ratios.

pricing kernel, heterogeneity

Abstract:
At the end of 1982 Denmark implemented a fiscal stabilization program which led to an economic expansion. This expansionary effect, as Giavazzi and Pagano (1990) and Bertola and Drazen (1993) argue, should be expected if consumers are forward- looking and the stabilization program is anticipated. This paper asks whether a similar model of forward-looking consumers is also capable of explaining the behavior of asset prices around the time of the stabilization. We argue that the behavior of Danish financial markets points in the direction of two interesting features of the policy change. First, in a model of intertemporal consumption smoothing, the observed stock market rally suggests that investors expected an increase in after-tax dividends. The term-structure evidence, on the other hand, is consistent with less than full credibility of the retrenchment, that is with investors attaching some probability to a further expansion of the government sector.

Abstract:
We consider: i) a dynamic investor who rebalances over time, treats returns as i.i.d., and accounts for learning, and ii) a static buy-and-hold investor who is aware of predictability and estimation risk. Both investors are internationally diversified and combine information on long- and short-history markets using cross-inference. For a dynamic investor, cross-inference generates separate hedging demands, and learning may generate positive hedging demands. For a static investor, some risky-asset allocations may decrease when estimation risk is ignored. Ignoring cross-inference, learning, and estimation risk, can generate sizable utility costs. Optimal investment in emerging markets, substantial in 1994, essentially drops to zero in 2000.

Abstract:
The risk premia assigned to economic (non-traded) risk factors can be decomposed into three parts: i) the risk premia on maximum-correlation portfolios mimicking the factors; ii) (minus) the covariance between the non-traded components of the candidate pricing kernel of a given model and the factors; and iii) (minus) the mis-pricing assigned by the candidate pricing kernel to the maximum-correlation mimicking portfolios. The first component is the same across asset-pricing models, and is typically estimated with little (absolute) bias and high precision. The second component, on the other hand, is essentially arbitrary, and can be estimated with large (absolute) biases and low precisions by multi-beta models with non-traded factors. This second component is also sensitive to the criterion minimized in estimation. The third component is estimated reasonably well, both for models with traded and non-traded factors. We conclude that the economic risk premia assigned by multi-beta models with non-traded factors can be very unreliable. Conversely, the risk premia on maximum-correlation portfolios provide more reliable indications of whether a non-traded risk factor is priced. These results hold for both the constant and the time-varying components of the factor risk premia.

economic risk premia, non-traded factors, maximum-correlation portfolios

Abstract:
The risk premia assigned to economic (non-traded) risk factors can be decomposed into three parts: i) the risk premia on maximum-correlation portfolios mimicking the factors; ii) (minus) the covariance between the non-traded components of the candidate pricing kernel of a given model and the factors; and iii) (minus) the mis-pricing assigned by the candidate pricing kernel to the maximum-correlation mimicking portfolios. The first component is the same across asset-pricing models, and is typically estimated with little (absolute) bias and high precision. The second component, on the other hand, is essentially arbitrary, and can be estimated with large (absolute) biases and low precisions by multi-beta models with non-traded factors. This second component is also sensitive to the criterion minimized in estimation. The third component is estimated reasonably well, both for models with traded and non-traded factors. We conclude that the economic risk premia assigned by multi-beta models with non-traded factors can be very unreliable. Conversely, the risk premia on maximum-correlation portfolios provide more reliable indications of whether a non-traded risk factor is priced. These results hold for both the constant and the time-varying components of the factor risk premia.

economic risk premia, non-traded factors, maximum-correlation portfolios

Abstract:
This paper considers two alternative formulations of the linear factor model (LFM) with nontraded factors. The first formulation is the traditional LFM, where the estimation of risk premia and alphas is performed by means of a cross-sectional regression of average returns on betas. The second formulation (LFM*) replaces the factors with their projections on the span of excess returns. This formulation requires only time-series regressions for the estimation of risk premia and alphas. We compare the theoretical properties of the two approaches and study the small-sample properties of estimates and test statistics. Our results show that when estimating risk premia and testing multi-beta models, the LFM* formulation should be considered in addition to, or even instead of, the more traditional LFM formulation.

mimicking portfolios, economic risk premia, multi-beta models

Abstract:
This paper develops a continuous-time version of the Lucas and Stokey (1987) cash-in-advance economy. The continuous-time setting allows for an exact characterization of equilibrium expected excess returns, real and nominal interest rates, the price level, and the inflation risk premium, for general preferences and stochastic processes. Explicit general-equilibrium solutions are obtained for separable logarithmic preferences and serially correlated processes for output and money growth. A characteristic feature of the economy is that even when preferences are separable in cash and credit goods, money supply still affects asset prices.

cash-in-advance constraint

Abstract:
We consider two formulations of the linear factor model with non-traded factors. In the first formulation (LFM), risk premia and alphas are estimated by a cross-sectional regression of average returns on betas. In the second formulation (LFM*), the factors are replaced by their projections on the span of excess returns, and risk premia and alphas are estimated by time-series regressions. We compare the two formulations and study the small-sample properties of estimates and test statistics. We conclude that the LFM* formulation should be considered in addition to, or even instead of, the more traditional LFM formulation.

Abstract:
The risk premia of linear factor models on economic (non-traded) risk factors can be decomposed into: i) the premium on maximum-correlation portfolios mimicking the factors; ii) (minus) the covariance between the non-traded components of the pricing kernel and the factors; and iii) (minus) the mispricing of the maximum-correlation portfolios. The first component is independent of a model's non-traded variability and mispricing, and is typically estimated with little bias and high precision. We conclude that the premia on maximum-correlation portfolios are appealing alternatives to the risk premia of linear factor models, with the dividend yield being the only economic factor significantly priced.

economic factors, risk premia, pricing kernel, maximum-correlation portfolio

Abstract:
The risk premia of linear factor models on economic (non-traded) risk factors can be decomposed into: i) the premium on maximum-correlation portfolios mimicking the factors; ii) (minus) the covariance between the non-traded components of the pricing kernel and the factors; and iii) (minus) the mispricing of the maximum-correlation portfolios. For a given set of assets available for investment, the first component is the same across models and is typically estimated with little bias and high precision. We conclude that the premia on maximum-correlation portfolios are appealing alternatives to the risk premia of linear factor models, with the dividend yield being the only economic factor significantly priced.

economic factors, risk premia, pricing kernel, maximum-correlation portfolio

Abstract:
We assume the short-term rate to revert towards a central tendency which in, turn, is stochastically changing over time. We impose minimal restrictions on the joint behavior of the short-term rate and the central-tendency factor, and derive implications for the term structure of interest rates. The analysis suggests a proxy for the central tendency which is then used to estimate the short-term rate process. Our model captures variations in the short-term rate better than the Vasicek (1977) and Cox, Ingersoll and Ross (1985) models, where the central tendency is assumed to be constant. Also, the central-tendency proxy explains the conditional volatility of the short-term rate better than the short-term rate itself.

term structure

Abstract:
We assume that the instantaneous riskless rate reverts towards a central tendency which, in turn, is changing stochastically over time, and we derive a model of the term structure of interest rates. Our term-structure model implies that a linear combination of any two rates can be used as a proxy of the central tendency. Based on the central-tendency proxy, we estimate a model of the one-month rate which performs better than models which assume the central tendency to be constant.

term structure

Abstract:
We assume that the instantaneous riskless rate reverts toward a central tendency which, in turn, is changing stochastically over time. As a result, current short-term rates are not sufficient to predict future short-term rates movements, as it would be the case if the central tendency was constant. However, since longer-maturity bond prices incorporate information about the central tendency, longer-maturity bond yields can be used to predict future short-term rate movements. We develop a two-factor model of the term-structure which implies that a linear combination of any two rates can be used as a proxy for the central tendency. Based on this central-tendency proxy, we estimate a model of the one-month rate which performs better than models which assume the central tendency to be constant.

term structure

Abstract:
We assume that the instantaneous riskless rate reverts toward a central tendency which, in turn, is changing stochastically over time. As a result, current short-term rates are not sufficient to predict future short-term rates movements, as it would be the case if the central tendency was constant. However, since longer-maturity bond prices incorporate information about the central tendency, longer-maturity bond yields can be used to predict future short-term rate movements. We develop a two-factor model of the term-structure which implies that a linear combination of any two rates can be used as a proxy for the central tendency. Based on this central-tendency proxy, we estimate a model of the one-month rate which performs better than models which assume the central tendency to be constant.

term structure

Abstract:
We assume that the instantaneous riskless rate reverts towards a central tendency which in turn, is changing stochastically over time. As a result, current short-term rates are not" sufficient to predict future short-term rates movements, as would be the case if the central" tendency was constant. However, since longer-maturity bond prices incorporate information" about the central tendency, longer-maturity bond yields can be used to predict future short-term" rate movements. We develop a two-factor model of the term-structure which implies that a" linear combination of any two rates can be used as a proxy for the central tendency. Based on" this central-tendency proxy, we estimate a model of the one-month rate which performs better" than models which assume the central tendency to be constant.

Abstract:
This paper examines newly-available intraday data from the interdealer government bond market to investigate the effects of economic-news announcements on prices, volume, and bid-ask spreads. By using expectational data we are able to separate out the impact of concurrent announcements. The use of intraday price data together with data on market expectations allows us to obtain new and different results relative to previous studies. We find several economic announcements to have a significant impact on two, ten, and thirty-year bond prices. For announcements that have a significant impact on bond prices, the impact occurs within one minute after the announcement. The three-month T-Bill price, on the other hand, is not impacted by the major economic announcement releases. This suggests that at least two factors of uncertainty are needed to model bond prices. For the two, ten, and thirty-year bonds we find a strong association between announcements and volume. Macroeconomic announcements do not have as much of an effect on trading volume for the three-month Treasury bill, although changes in monetary policy lead to an average trading volume up to nine times higher than at non-announcement times. Immediately after most economic announcements, bid-asl spreads widen significantly, while they tighten in the next five to fifteen minutes

Abstract:
This paper examines newly-available intra-day data from the inter-dealer government bond market to investigate the effects of economic-news announcements on prices, trading volume, and bid-ask spreads. The use of intra-day price data together with data on market expectations allows us to obtain new and different results relative to previous studies. We find a total of seventeen economic announcements to have a significant impact on the price of at least one of the following instruments: a three-month bill, a two ' and ten year note, and a thirty year bond. Ten of them significantly affect all note and bond prices. For announcements that have a significant impact on prices, the impact occurs within one minute after the announcement. Interestingly, only three announcements affect the bill price. This suggests that at least two factors of uncertainty are needed to model the yield curve. For the ten-year note we find a strong association between announcements and trading volume. Economic announcements have less effect on trading volume for the three-month bill, although changes in monetary policy lead to an average trading volume up to nine times higher than at non-announcement times. Bid-ask spreads widen immediately after most economic announcements, but then return to normal levels within 5 to 15 minutes. For almost all announcements, volatility is significantly higher after the release, especially for the announcements that significantly affect prices.

Abstract:
This paper reexamines the proxy hypothesis of Fama (American Economic Review, 1981, 71, 545-565) as the main explanation for the negative correlation between stock returns and inflation. We look at quarterly data on industrial-production growth, monetary-base growth, CPI inflation, three-month Treasury-bill rates, and returns on the equally-weighted NYSE portfolio, for the 1954-1976 and 1977-1990 periods. Using time-series techniques, we find that production growth induces only a weak negative correlation between inflation and stock returns, and explains less of the covariance between the two series than inflation and interest-rate innovations.

vector autoregression, vector moving average, covariance decomposition

Abstract:
We consider the impact of transaction costs on the portfolio decisions of a long-lived agent with isoelastic preferences. In particular, we focus on how portfolio choice, rebalancing frequency and average cost incurred change over the lifecycle are affected by return predictability. Two types of costs are evaluated: proportional to the change in the holding of the risky asset and a fixed fraction of portfolio value. We find that realistic transaction costs can materially affect rebalancing behavior, creating no-trade regions that widen near the investor's terminal date. At the same time, realistic proportional and fixed costs have little effect on the midpoint of the no-trade region, unless liquidation costs differ across assets. Return predictability calibrated to U.S. stock returns is found to have large effects on rebalancing behavior relative to independent and identically distributed (i.i.d.) returns with the same unconditional distribution. For example, return predictability causes rebalancing frequency to increase, and cost incurred to increase by an order of magnitude, at all points in the investor's life. No-trade regions early in life are wider when returns are predictable than when they are not. Finally, we find that the nature of the return predictability, including the presence or not of return heteroscedasticity, can have large effects on rebalancing behavior.

Abstract:
We explore the effects of overnight-rate targeting on nominal interest rates of longer maturities. In a realistic model of noisy targeting and infrequent target changes, expectations of future policy actions introduce persistent spreads between interest rates of different maturities. Some empirical features of U.S. money-market daily interest rate data are broadly consistent with our theoretical assumptions and results. Not surprisingly, however, the data reject the expectations-hypothesis (EH) relation that we take as a working assumptions. A newly available series of historical interest-rate targets and simple tests based on our theoretical insights suggest that the EH rejection may be due to erroneous market expectations of the policy-induced component of fed funds dynamics. We briefly discuss how the size and volatility of such expectations may be interpreted from the perspective of our theoretical framework.

monetary regimes, expectaions hypothesis, peso problem

Abstract:
No abstract is available for this paper.

Abstract:
Real money balances are held separately for consumption and portfolio reasons. When real balances are a state variable in the investor s optimization problem, there is a specific inflation-hedging portfolio. An investor hedges against inflation when the effect of real money holdings on the marginal utility of wealth is negative. We show that an increase in real balances due to inflation has two opposite effects on the marginal utility of wealth. On the one hand, the decrease in the real balances reduces consumption, which in turn raises the marginal utility and decreases the marginal cost of consuming: this explains why an investor would normally hedge inflation. One the other hand, the decrease in real balances tends to increase the marginal cost of consuming. When this second effect dominates, we have the somewhat surprising result that the investor reverse-hedges inflation.

Abstract:
We consider the implications for mean factor risk premia for the variance of admissible (normalized) stochastic discount factors, or pricing kernels. For given mean risk premia, we identify lower bounds on the variance of the pricing kernel which exceed the variance of the projection of the pricing kernel on the (augmented) asset return space: the â¬SHansen and Jagannathanâ¬? variance bound. These lower bounds increase with the covariability between the components of the pricing kernel and of the factors which are not explained by asset returns, and decrease with the distance between the factors and the (augmented) asset-return space. As an application, we show that the inflation risk premium generated by a consumption-based pricing kernel implies a standard deviation of the kernel which is up to 15% higher than the Hansen and Jagannathan bound.

Abstract:
In a variety of realistic scenarios, some investors trade infrequently rather than continuously, basing their buy or sell decisions on current price levels. A â¬Sprice barrierâ¬? is a price level at which a large number of investors either buy or sell securities. We analyze the dynamics of asset prices in an economy with infrequent traders and price barriers. Our analysis predicts that when price barriers exist, both asset prices and price volatility can jump at the time the price barrier is reached, even if the trade is rationally anticipated. Moreover, the direction of the price jump may very well be the opposite of what one would expect. A price-triggered purchase may generate a downward jump in stock prices and, vice versa, a price-triggered sale may induce stock prices to jump above the price barrier. This is because trades affect prices before they are implemented: the anticipation of a stock purchase inflates stock prices, while the anticipation of a stock sale depresses stock prices. In the case of repeated trades, before-trade prices anticipate not only the next trade, but also the following ones. This may lead to the counterintuitive result that stock prices are inflated rather than depressed, in the proximity of a stock sale.

Abstract:
This paper studies the effects on financial markets of an anticipated fiscal stabilization policy in a stochastic environment. Stabilization is defined as a discrete change in the budget process which is implemented when government consumption reaches some threshold level, known by economic agents. Our analysis integrates the study of financial markets within the framework adopted by Bertola and Drazen (1993) to explain the effects of private consumption of an anticipated fiscal retrenchment, such as the fiscal reform implemented in Denmark in 1983. The actual behavior of Danish financial markets points in the direction of two interesting features of the policy change. First, in a model intertemporal consumption smoothing, one can replicate the observed boom in the stock market only with expectations of an increase in net income to stock-holders. The term-structure evidence on the other hand, is consistent with less than full credibility of the retrenchment, that is with investors attaching some probability to a further expansion of the government sector.

fiscal retrenchments, government spending, policy regime, stock prices, term structure of interst rates, trigger point

Abstract:
We model an economy where stocks and bonds (consols) are traded by two types of agents: speculators, expected utility maximizers always present in the market, and infrequent traders, whose trading motives are not explicitly modeled. A solution technique for equilibrium prices is developed when trades are triggered by stock prices reaching some threshold level, corresponding to a specific value of the dividend flow. Across trade scenarios we find expectations of stock sales to depress stock prices relative to the no-trade case, while expectations of stock purchases tend to inflate them .both effects bring about heteroskedasticity and predictability of stock returns. Our analysis yields insights as to the equilibrium effects of a variety of trading strategies, which mechanically generate market orders in response to changes in stock prices.

Abstract:
This paper studies a newly available data set of daily net transfers in several asset classes for roughly 1.5 million 401(k) participants, during the August 1997-September 2002 period. The main stylized fact is that 401(k) transfers correlate strongly and positively with the same-day asset returns in the corresponding asset class. This result is important regardless of the direction of causality. If transfers cause returns, then being able to time the market based on 401(k) transfers would be very profitable. If returns cause transfers, then changing allocations, in response to same-day returns, can have a substantial negative effect on expected utility. Indeed, by using the identification-through-heteroskedasticity technique, we show that there is both an immediate reaction of transfers to returns and an immediate reaction of returns to transfers.

401(k) plans

Abstract:
This paper combines the continuous arrival of information with the infrequency of trades, and investigates the effects on asset price dynamics of positive- and negative-feedback trading. Specifically, we model an economy where stocks and bonds are traded by two types of agents: speculators who maximize expected utility, and feedback traders who mechanically respond to price changes and infrequently submit market orders. We show that positive-feedback strategies increase the volatility of stock returns, and the response of stock prices to dividend news. Conversely, the presence of negative-feedback traders makes stock returns less volatile, and prices less responsive to dividends.

Abstract:
If a pricing kernel assigns a premium to a risk variable that differs from the one assigned by the minimum-variance admissible kernel, then the pricing kernel must exhibit more variability than the minimum-variance kernel. Based on this intuition, we derive a variance bound that is more stringent than that of Hansen and Jagannathan (1991). When we apply our bound to the kernel of a representative consumer with power utility, we find that the consumption risk premium increases the severity of the "equity-premium puzzle" of Mehra and Prescott (1985).

Abstract:
This paper studies the implications of security-market data for the risk premia on economic risk variables and the variance of stochastic discount factors, or pricing kernels. We derive a lower bound on the variance of pricing kernels consistent with asset returns, given the risk premium they assign to an economic risk variable. The difference between our lower variance bound and the lower bound derived by Hansen and Jagannathan (1991) is equal to the squared Sharpe ratio of a portfolio long in the economic risk variable and short in its nearest hedge using available securities. Our variance bounds go one step further in studying the link between the pricing implications of a kernel and the minimum-variance restrictions that it needs to satisfy. We also derive bounds on the risk premia associated with economic risk variables, given apriori information on the variance of the stochastic discount factor. We find that for risk variables such as inflation and consumption growth our lower bound on the pricing-kernel variance is significantly more stringent than the Hansen and Jagannathan (1991) bound. We apply our analysis to the standard representative-agent consumption-based pricing kernel with isoelastic preferences, and we find that the implications of the risk premium on consumption growth exacerbate the "equity-premium puzzle" relative to previous studies.

Abstract:
A price barrier is a price level at which a large number of investors either buy or sell securities. We analyze the dynamics of asset prices in an economy in which price barriers exist. Our analysis suggests that asset prices and volatility can exhibit jumps when the price barrier is reached. Interestingly, the market's anticipation of future trades can influence prices in the opposite direction from what one might expect. For example, when multiple barriers exist, stock prices can be inflated, rather than depressed, in the proximity of an anticipated stock sale.

Abstract:
In a variety of realistic scenarios, some investors trade infrequently rather than continuously, basing their buy or sell decisions on current price levels. A "price barrier" is a price level at which a large number of investors either buy or sell securities. We analyze the dynamics of asset prices in an economy with infrequent traders and price barriers. Our analysis predicts that when price barriers exist, both asset prices and price volatility can jump at the time the price barrier is reached, even if the trade is rationally anticipated. Moreover, the direction of the price jump may very well be the opposite of what one would expect. A price-triggered purchase may generate a downward jump in stock prices and, vice versa, a price-triggered sale may induce stock prices to jump above the price barrier. This is because trades affect prices before they are implemented: the anticipation of a stock purchase inflates stock prices, while the anticipation of a stock sale depresses stock prices. In the case of repeated trades, before-trade prices anticipate not only the next trade, but also the following ones. This may lead to the counterintuitive result that stock prices are inflated, rather than depressed, in the proximity of a stock sale.